Addition of Non-Skin Classes in Skin Type Classification Using EfficientNet-B0 Architecture
DOI:
https://doi.org/10.30871/jaic.v9i5.10335Keywords:
Skin Classification, Nonskin Integration, Efficientnet-B0, Deep Learning, Image ClassificationAbstract
Skin type classification is an essential process in dermatology and skincare, aiming to categorize skin conditions such as dry, normal, and oily. However, image-based skin classification models often struggle when confronted with non-skin objects like clothing, background, or hair that are not accounted for in standard datasets. This study proposes a novel approach by integrating a nonskin class into a skin type classification model based on the EfficientNet-B0 architecture. The dataset used consists of images categorized into four classes: dry, normal, oily, and nonskin. The model was trained using transfer learning and optimized through techniques such as data augmentation, learning rate scheduling, and early stopping. The final evaluation achieved an accuracy of 91%, with the nonskin class showing perfect precision and recall. These results demonstrate that incorporating nonskin data can significantly enhance model robustness and accuracy. This research contributes a practical method for improving the reliability of skin classification systems, especially in real-world environments.
Downloads
References
[1] R. Oliveira, J. Ferreira, L. F. Azevedo, and I. F. Almeida, “An Overview of Methods to Characterize Skin Type: Focus on Visual Rating Scales and Self-Report Instruments,” Cosmetics, vol. 10, no. 1, p. 14, Jan. 2023, doi: 10.3390/cosmetics10010014.
[2] S. Saidah, Y. N. Fuadah, F. Alia, N. Ibrahim, R. Magdalena, and S. Rizal, “Facial Skin Type Classification Based on Microscopic Images Using Convolutional Neural Network (CNN),” 2021, pp. 75–83. doi: 10.1007/978-981-33-6926-9_7.
[3] S. Saiwaeo, S. Arwatchananukul, L. Mungmai, W. Preedalikit, and N. Aunsri, “Human skin type classification using image processing and deep learning approaches,” Heliyon, vol. 9, no. 11, p. e21176, 2023, doi: 10.1016/j.heliyon.2023.e21176.
[4] O. Iparraguirre-Villanueva and M. Cabanillas-Carbonell, “Application of Convolutional Neural Networks in Skin Disease Prediction: Accuracy and Efficiency in Dermatological Image Analysis,” International Journal of Online and Biomedical Engineering (iJOE), vol. 21, no. 02, pp. 18–37, Feb. 2025, doi: 10.3991/ijoe.v21i02.52871.
[5] S. Inthiyaz et al., “Skin disease detection using deep learning,” Advances in Engineering Software, vol. 175, p. 103361, Jan. 2023, doi: 10.1016/j.advengsoft.2022.103361.
[6] M. Y. Shams et al., “Skin Disease Classification: A Comparison of ResNet50, MobileNet, and Efficient-B0,” Journal of Current Multidisciplinary Research, vol. 1, no. 1, pp. 1–7, Jan. 2025, doi: 10.21608/jcmr.2025.327880.1002.
[7] H. Van Thanh, N. D. Quang, T. M. Phuong, K.-H. Jo, and V.-D. Hoang, “A compact version of EfficientNet for skin disease diagnosis application,” Neurocomputing, vol. 620, p. 129166, Mar. 2025, doi: 10.1016/j.neucom.2024.129166.
[8] M. Harahap et al., “Skin cancer classification using EfficientNet architecture,” Bulletin of Electrical Engineering and Informatics, vol. 13, no. 4, pp. 2716–2728, Aug. 2024, doi: 10.11591/eei.v13i4.7159.
[9] P. Khairunnisa, W. E. Putra, W. Yitong, A. Jufrizal, and M. N. A. Makmum, “Convolutional Neural Networks Using EfficientNetB0 Architecture and Hyperparameters on Skin Disease Classification,” Public Research Journal of Engineering, Data Technology and Computer Science, vol. 2, no. 1, pp. 127–137, 2025.
[10] T. Talaei Khoei, H. Ould Slimane, and N. Kaabouch, “Deep learning: systematic review, models, challenges, and research directions,” Neural Comput Appl, vol. 35, no. 31, pp. 23103–23124, Nov. 2023, doi: 10.1007/s00521-023-08957-4.
[11] A. Xie, X. Zhang, Q. Huang, and J. Wu, “Utilizing Untargeted Lipidomics Technology to Elucidate Differences in Lipid Compositions Among Sensitive Dry, Sensitive Oily and Healthy Skin Types,” Metabolites, vol. 15, no. 5, p. 292, Apr. 2025, doi: 10.3390/metabo15050292.
[12] M. Tan and Q. V Le, “EfficientNet : Rethinking Model Scaling for Convolutional Neural Networks,” 2019.
[13] P. Zhang and D. Chaudhary, “Hybrid Deep Learning Framework for Enhanced Melanoma Detection,” Jul. 2024, [Online]. Available: http://arxiv.org/abs/2408.00772
[14] N. Alipour, T. Burke, and J. Courtney, “Skin Type Diversity in Skin Lesion Datasets: A Review,” Curr Dermatol Rep, vol. 13, no. 3, pp. 198–210, Aug. 2024, doi: 10.1007/s13671-024-00440-0.
[15] K. Ramamurthy, I. Thayumanaswamy, M. Radhakrishnan, D. Won, and S. Lingaswamy, “Integration of Localized, Contextual, and Hierarchical Features in Deep Learning for Improved Skin Lesion Classification,” Diagnostics, vol. 14, no. 13, p. 1338, Jun. 2024, doi: 10.3390/diagnostics14131338.
[16] J. Raihan and S. A. Ridita, “PhytoCare : A hybrid approach for identifying Rice , Potato and Corn diseases PhytoCare : A hybrid approach for identifying Rice , Potato and Corn diseases,” no. February, 2024.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Haitsam Muftin Sani, Ajie Kusuma Wardhana
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) ) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
